Diagnostic stability in individuals with autism spectrum disorder: insights from a longitudinal follow-up study.

Researchers followed 155 people from age 2 to 25 to see how autism diagnoses change over time. They found that some people gained or lost autism diagnoses as they grew up. For children with average intelligence who eventually lost their autism diagnosis, early signs of this change could be seen by age 5. The study shows that autism diagnoses aren't always permanent and can change as people develop, especially for those without intellectual disability.

This longitudinal study followed 155 individuals from age 2 to 25 years to examine how autism diagnoses change over time. Participants were assessed at multiple time points with blinded assessors and categorized into four diagnostic trajectory groups: Retained ASD, Lost ASD, Never Had ASD, or Gained ASD. Results showed that autism diagnoses can shift across development, with gradual changes in core autism symptoms beginning in childhood. For individuals with average intelligence (VIQ > 70), those who eventually lost their diagnosis could be distinguished from those who retained it as early as age 5.

Changes were more complex for those with intellectual disability, where differential diagnosis was challenging.

Results suggest autism diagnoses should be viewed developmentally rather than as fixed labels. Clinicians should consider ongoing assessment across development, particularly for individuals with average intelligence. Early symptom patterns may help predict diagnostic trajectories.

Single cohort study design limits generalizability. Complex interactions between autism symptoms and intellectual disability make interpretation challenging for individuals with VIQ < 70. The study doesn't specify causes of diagnostic changes or interventions received.

This longitudinal study of autism symptom trajectories provides unique information that can characterize autism features and diagnostic patterns from childhood to adulthood. Participants (n = 155) were part of a longitudinal cohort referred for possible autism where in-person assessments were completed at ages 2, 3, 5, 9, 19, and 25. Assessors were blinded to previous diagnoses. Based on adult best estimate diagnoses, participants were categorized into one of the four groups: Retained ASD, Lost ASD, Never Had ASD, or Gained ASD Diagnosis.

To examine developmental changes in autism symptoms, mixed models indicated the rate of change in ADOS CSS and ADI-R scores in each diagnostic group. A subset of participants with VIQ> and <70 were assigned a diagnosis in adulthood that differed from diagnoses earlier in development. Across cognitive levels, the majority of novel diagnoses emerged in adulthood. For those with VIQ > 70, improvements in ADOS CSS over time for the Lost Diagnosis group and worsening in CSS in the Gained Diagnosis group were gradual.

Individuals with VIQ > 70 who lost a diagnosis even in adulthood could be distinguished on CSS and ADI-R scores by age 5 from those who retained their ASD diagnosis. Although most participants with VIQ < 70 saw decreases in autistic symptoms as a whole, changes in autism diagnoses were confounded by disentangling profound intellectual disability as a differential diagnosis or co-occurrence. Only the Never Had Diagnosis group revealed significant changes in ADOS scores over time, with autism symptoms increasing. Associated with gradual changes in core features of autism beginning in childhood, diagnoses of autism can shift across development.